Linkage mapping of Quantitative Trait Loci (QTL) in inbred line crosses of animal (mouse) models of human diseases and association mapping of QTL in human populations for genome- wide gene expression traits and disease phenotypes (and other 'omics', e.g. metabolomics, traits) are very powerful approaches to the identification of genes, pathways and regulatory networks underlying complex human diseases. Most of these systems genetics analyses are performed in multiple steps, with QTL mapping on all variables of interest (e.g., gene expression profiles, disease related phenotypes) being a very critical first step. For this first step, methods for standard QTL mapping have been used, i.e. methods for QTL mapping on a small number of traits, and even more recently developed methods for QTL mapping in systems genetics still do not fully account for the very highly multivariate structure of the traits and are hence underpowered. In this proposal our goal is to implement, evaluate and compare several types of multivariate analyses for very highly multivariate QTL mapping in systems genetics, to understand the relationships between these multivariate methods, and to determine how to best perform highly multivariate QTL linkage and association mapping in order to gain the most power and accuracy. Our main focus will be on multivariate methods for exploring associations between two groups of variables (marker genotype profiles versus disease phenotypes and gene expression profiles in a segregating population), but we will also consider extensions to more than two sets of variables (such as marker genotype profiles, gene expression profiles, and DNA methylation profiles). We will focus on methods based on sparse Bayesian and frequentist canonical correlation analysis, sparse partial least-squares analysis and biclustering. These methods will be evaluated with simulated and real systems genetics data. The successful methods will be implemented in C++ code for efficiency and made available as an R package for user-friendliness.

Public Health Relevance

Linkage mapping of Quantitative Trait Loci (QTL) in inbred line crosses of animal (mouse) models of human diseases and association mapping of QTL in human populations for genome- wide gene expression traits and disease phenotypes (and other 'omics', e.g. metabolomics, traits) are critical steps in systems genetics approaches to the identification of genes, pathways and regulatory networks underlying complex human diseases, such as cardiovascular disease, obesity, osteoporosis, arthritis and so on. Identifying key targets in regulatory networks underlying diseases is a critical step towards the development of better therapies and preventive measures for these complex diseases.

Agency
National Institute of Health (NIH)
Institute
National Human Genome Research Institute (NHGRI)
Type
Research Project (R01)
Project #
5R01HG005254-03
Application #
8197687
Study Section
Genomics, Computational Biology and Technology Study Section (GCAT)
Program Officer
Brooks, Lisa
Project Start
2010-01-28
Project End
2014-11-30
Budget Start
2011-12-01
Budget End
2014-11-30
Support Year
3
Fiscal Year
2012
Total Cost
$268,028
Indirect Cost
$94,778
Name
Virginia Polytechnic Institute and State University
Department
Type
Organized Research Units
DUNS #
003137015
City
Blacksburg
State
VA
Country
United States
Zip Code
24061
Yi, Hui; Breheny, Patrick; Imam, Netsanet et al. (2015) Penalized multimarker vs. single-marker regression methods for genome-wide association studies of quantitative traits. Genetics 199:205-22
Pinna, Andrea; Soranzo, Nicola; Hoeschele, Ina et al. (2011) Simulating systems genetics data with SysGenSIM. Bioinformatics 27:2459-62